Overclocking the Venice Core
For our first test comparing the Newcastle and Venice cores, we slowly raised each processor's HT clock speed via our motherboard's BIOS until the test system was no longer stable. Please note that each processor was running at its stock voltage in this test, 1.4v for the Venice based CPU and 1.5v for the Newcastle. The heatsink used to cool each processor was an all-copper model from Thermaltake, the Silent Boost K8, equipped with an 80mm, 2500RPM fan. This heatsink is fairly large, and has thin copper fins over its entire surface.
To be certain the CPU was the limiting factor in our overclocking attempts, we dropped the HyperTransport clock multiplier to 3X (600MHz), from its default of 5X (1GHz), locked the PCI Express clock to 100MHz, and ran our memory at timings of 2.5-3-3-10 at 2.8v. At these settings, this particular configuration topped out at an HT clock of over 280MHz in a previous article...
With both of our Athlon 64 3800+ processors operating with their default multiplier of 12, and with our Gigabyte motherboard configured for a 200MHz HT clock, both processors ran at a "default" clock speed of 2412MHz; 12MHz above the expected speed of 2400MHz. To begin the experiment, we then raised the HT clock speed via the motherboard's BIOS menu to see which CPU would top out first. In the end, the .09 micron, Venice core based 3800+ hit a maximum clock speed of 2640MHz (12 x 220MHz), besting the .13 micron Newcastle by only 58MHz. At default settings, our particular Venice core based A64 seemed a bit more "overclocker friendly", but obviously the clock speed difference wasn't terribly substantial.
Now, before you come to the conclusion that we've inhaled the fumes from one too many fried PSUs, please be aware that there are a number of other measures we could have taken to achieve even higher clock speeds in this type of test. And yes, we've seen the reports of Venice cores easily reaching 2.8GHz on air, but remember this test was conducted with the CPU at its stock voltage. Our goal was to find each processor's peak clock speed, without tweaking the system in any way.
When we did goose the voltage a bit, up to 1.6v in this case, our particular Venice core based Athon 64 3800+ had no problem hitting 2.77GHz. With some more exotic cooling, and maybe a little more voltage, we wouldn't be surprised if this particular CPU could hit 2.85GHz, or even 2.9GHz. Perhaps we'll have to experiment with a VapoChill in a future article? 3GHz+ could be in this processor's future. :-)